(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "A PIC compatible RISC CPU core Implementation for FPGA based Configurable SOC platform for Embedded Applications"

ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 201 1 



A PIC compatible RISC CPU core Implementation for 

FPGA based Configurable SOC platform for 

Embedded Applications 

Haritha Madasi ', S Nagakishore Bhavanam 2 , Vaujadevi Midasala 3 

1 Holy Mary Institute of Technology & Science, E.C.EDept., Hyderabad, India 

Email: harikha_bujji@yahoo.com 

2 Guru Nanak Institute of Science & Technology, E.C.E Dept., Hyderabad, India 

Email: kishorereddy.vlsi@gmail.com 

3 Keshav Memorial Institute of Technology, E.C.E Dept., Hyderabad, India 

Email: vasujadevi@gmail.com 



Abstract — Modern embedded systems are built around the soft 
core processors implemented on FPGA. The FPGAs being 
capable of implementing custom hardware blocks giving the 
advantage of ASICs, and allowing the implementation of 
processor platform are resulting in powerful Configurable- 
system on chip(C-SoC)platforms. The Microchip's PIC 
microcontroller is very widely used microcontroller 
architecture across various embedded systems. The 
implementation of such core on FPGA is very much useful in 
CSOC based embedded systems. This type of designs can be 
widely used in those controlling fields demanding low power 
consumption and high ratio of performance to price. In this 
project a reduced instruction set computer (RISC) CPU IP 
core whose instructions are compatible with the Microchip 
PIC16C6Xseries of microcontrollers is implemented in VHDL. 
The core is based on 8-bit RISC architecture and top-Down 
design methodology is used in developing the core. The RISC 
CPU core is based on Harvard architecture with 14-bit 
instruction length and 8-bit data length and two-stage 
instruction pipeline. The architecture will be designed aiming 
at single cycle execution of the instructions, except those 
related to program branches. Since this type of CPU based on 
RISC architecture, there are only 35 reduced instructions in 
its instruction set, which are easy to be learned and used. The 
performance of the 8-bit RISC CPU is better than those of 
CPUs which are based on CISC architecture. Modelsim Xilinx 
Edition (MXE) will be used simulation and functional 
verification. The Xilinx Spartan-3E FPGAs will be used 
synthesis and timing analysis. The results will be verified on 
chip with chipscope tool. 

Keywords- CPU; RISC; IP Core; IC; MXE; Xilix. 

I. Pic Microcontrollers 

PIC microcontrollers are finding their way into new 
applications like smart phones, audio accessories, video 
gaming peripherals and advanced medical devices. 
Microchip provides solutions for the entire performance 
range of 8-bit microcontrollers, with easy-to-use development 
tools, complete technical documentation and post design- 
in support through a global sales and distribution network. 
There are hundreds of 8-bit PIC microcontrollers to choose 
from ranging from 6 to 100 pins and up to 128 KB Flash that 

©2011 ACEEE 
DOI:01.DEPE.02.02.44 



11 



are pin and code compatible across the portfolio. PIC 
microcontrollers with XLP technology feature the world's 
lowest active and sleep power consumption with flexible 
power modes and wake-up sources. MPLAB® Integrate 
development Environ-ment supports all PIC microcontrollers 
with C Compiler support and common development boards. 
Peripheral integration is key with communication and control 
peripherals like SPI, I 2 C™, UART, PWM, ADC and comparators as 
well as specialized peripherals for USB, mTouch™ Sensing, 
LCD, CAN and Ethernet. Customers have made PIC MCUs a 
worldwide standard, with over one million development 
systems shipped. PIC microcontrollers are quick and easy 
to design into a wide variety of applications with a long history 
of dependable product deliveryfl]. 

H. INTRODUCTION TO PIC 16X MICROCONTROLLER 

The PIC16C63A/65B/73B/74B devices are low-cost, 
high-performance, CMOS, fully-static, 8 bit microcontrollers 
in the PIC16CXX mid-range family. All PIC micro® 
microcontrollers employ an advanced RISC architecture[2]. 
The PIC16CXX microcontroller family has enhanced core 
features, eight-level deep stack and multiple internal and 
external interrupt sources. The separate instruction and 
data buses of the Harvard architecture allow a 14-bit wide 
instruction word with the separate 8-bit wide data. The two 
stage instruction pipeline allows all instructions to execute 
in a single cycle, except for program branches, which require 
two cycles. A total of 35 instructions (reduced instruction 
set) are available. Additionally, a large register set gives 
some of the architectural innovations used to achieve a very 
high performance[l]. The PIC16C63A/73B devices have 22 
I/O pins. ThePIC16C65B/74B devices have 33 I/O pins. 
Each device has 192 bytes of RAM. In addition, several 
peripheral features are available including: three timer/ 
counters, two Capture/Compare/PWM modules and two 
serial ports. The Synchronous Serial Port (SSP) can be 
configured as either a 3-wire Serial Peripheral Interface (SPI) 
or the two-wire Inter-Integrated Circuit(I^C) bus. The 
Universal Synchronous Asynchronous Receiver 
Transmitter (USART) is also known as the Serial 
Communications Interface or SCI. Also, a 5- channel high- 



?VACEEE 



ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 201 1 



speed 8-bit A/D is provided on the PIC16C73B, while 
the PIC16C74B offers 8 channels. The 8-bit resolution is 
ideally suited for applications requiring low-cost analog 
interface, e.g., thermostat control, pressure sensing, etc. 
The PIC16C63A/65B/73B/74B devices have special 
features to reduce external components, thus reducing cost, 
enhancing system reliability and reducing power 
consumption. There are four oscillator options, of which 
the single pin RC oscillator provides a low-cost solution, 
the LP oscillator minimizes power consumption, XT is a 
standard crystal, and the HS is for high speed crystals. The 
SLEEP (power-down) feature provides a power- saving mode. 
The user can wake up the chip from SLEEP through several 
external and internal interrupts and resets. A highly reliable 
Watchdog Timer (WDT), with its own on-chip RC oscillator, 
provides protection against soft- ware lock-up, and also 
provides one way of waking the device from SLEEP. A UV 
erasable CERDIP packaged version is ideal for code 
development, while the cost-effective One-Time- 
Programmable (OTP) version is suitable for production in any 
volume. The PIC16C63A/65B/73B/74B devices fit nicely in 
many applications ranging from security and remote sensors 
to appliance control and automotive[l]. The EPROM 
technology makes customization of application programs 
(transmitter codes, motor speeds, receiver frequencies, etc.) 
extremely fast and convenient. The small footprint packages 
make this microcontroller series perfect for all applications with 
space limitations. Low cost, low power, high performance, ease 
of use and I/O flexibility make the PIC16C63 A/65B/73B/74B 
devices very versatile, even in areas where no microcontroller 
use has been considered before (e.g., timer functions, serial 
communication, capture and compare, PWM functions and 
coprocessor applications)[3]. 

IE. RISC IP DESIGN 

The CPU IP Core is an instruction-compatible of the 
Microchip PIC16C6X -series microcontrollers. The CPU 
having 22bit I/O ports only has 35 instructions which make 
the final hardware implementation compact. The CPU can 
address up to 2048 instruction words. And we use two level 
pipe-line structures to improve the executing speed. Figure 
3-1 below shows the main modules of the RISC CPU IP 
Core[2]. 




Figure 3-1. Main modules of RISC CPU IP Core 



As figure 3-1 show, the CPU module contains the datapath 
and controller modules while the program store is external to 
the CPU Core. This means that the Program Store can be 
located off-chip if needed. Locating the Program Store on- 
chip however enables the CPU Core to execute at the internal 
clock speed of CPU. 






F^Jtfl 



XlH 



Pngar 
I ■ 



K*H 






1*1 1 



^vaTO* 



| Mm.fr.frwg | 



i 



/ Ha*ur\ 

XT- 



i f* ■*; j C= 



i-iiu-:' 

..-.■. l •:, '■■ 

XI I 



moom 



H« <&'&<» 



T-Pf 



>.'","- m 






w 






^ 







■TOT vs.vn 



«xm 



J 



■'••■ 



-■:■■ 

IB 

Hca 

HI 

n 

as 

=■;: 



Figure 3-2. CPU architecture 

Figure 3-2 shows the CPU architectural overview. To begin 
with, the CPU uses Harvard architecture, in which, program 
and data are accessed from separate memories using separate 
buses. This improves bandwidth over traditional von eumann 
architecture where program and data may be fetched from 
the same memory using the same bus. Separating program 
and data busses further allows instructions to be sized 
differently than 8-bit wide data words. Instruction words are 
14-bits wide making it possible to have all single word 
instructions. A 14-bit wide program memory access bus 
fetches a 14-bit instruction in a single cycle. A two-stage 
pipeline overlaps fetch and execution of instructions. 
Consequently, all instructions execute in a single cycle except 
for program branches. The CPU can directly or indirectly 
address its register files or data memory. All special function 
registers (including the program counter) are mapped in the 
data memory. The CPU has a symmetrical instruction set that 
makes it possible to carry out any operation on any register 
using any addressing mode. This symmetrical nature makes 
programming with the CPU simple yet efficient, thus 
significantly reducing the learning curve. The CPU contains 
an 8-bit ALU and working register (W). The ALU is a general 
arithmetic unit. It performs arithmetic and Boolean functions 
between data in the working register and any register file. The 
ALU is 8-bits wide and capable of addition, subtraction, shift, 
and logical operations. Unless otherwise mentioned, arithmetic 
operations are two's complement in nature. In two-operand 
instructions, typically one operand is the working register (W 
register), the other operand is a file register or an immediate 
constant. In single operand instructions, the operand is either 
the working register or a file register. 



©2011 ACEEE 
DOI:01.IJEPE02.02.44 



12 



^ACEEE 



ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 201 1 



fcrt_»_ni 




-i lriB?.d»n Ri^itti 



/ir Vr^% 



/ *a*Ajwhiv\ 



RAMft|.jrt!Kl (j 



Figure 3-3. Simplified datapath 

Figure 3-3 shows a somewhat simplified datapath 
architecture where the eight Special Function Registers are 
not shown. This datapath is sufficient to execute most 
instructions so it will be used to introduce the overall CPU IP 
Core architecture. Most instructions in the datapath 
architecture operate on either the 'W or one of the 'f 
registers. The 'W register is input as the ALU's first operand 
while the currently addressed 'f register is input as the 
second ALU operand. The ALU result is output on 'alu_bus' 
from where it is loaded into either the 'W register or into the 
currently addressed 'f ' register. The Controller either asserts 
the 'load_w_reg' or the 'load_ram_reg' depending on if the 
instruction word in the instruction Register directs the result 
to the 'W or 'f register respectively [4] . In order for the ALU 
to be able to update the STATUS register bits Z, C, and DC a 
separate 'status_bus' exists together with three control 
signals named 'load_z', 'load_d' and 'load_dc'. These signals 
are displayed below in figure 4. Also note that figure 4 
contains the 'ALU Mux' on operand input two which allows 
the Controller to select the second ALU operand to come 
from either the RAM Registers (shown above in figure 3) or 
from one of the eight SFR (Special Function Register) 
Registers [0..7]. The ALU knows from the 'ir_reg_bus' which 
instruction word is currently being executed. The 'carry_bit' 
input is connected to the C bit in the STATUS register. This 
is used during additions and subtractions. The 
'ram_data_bus' input to the ALU Mux comes from the RAM 
registers as shown above in figure 3-4. The 'sfr_data_bus' 
comes from the currently addressed SFR register. 






i^.«ti-/i 



a 












fi 



J 



WRtpM' 



Figure 3-4. Datapath surrounding ALU 

The CPU architecture allows both direct and indirect 
addresses. A direct address is encoded directly in the 14-bit 

©2011 ACEEE 
DOI:01.DEPE.02.02.44 



instruction word while an indirect address is taken from the 
FSR Special Function Register. The address generation logic 
is shown below in figure3-5. 



* ■. • ■ • 




I 



SfRRtfttr 



»J» fi 



7 



V r t 7 

*■—■-■' \ Pistil AddrrulAii 

V 



-I'-.-l 



L 



E 



TT 



ftJUU 



i J"^J^^_ 



0**_**«»*Jt* 



Figure 3-5. Direct and indirect address selection 

As can be seen in Figure 3-5, the Register Address 
Multiplexer passes through either the 'fffffff field in the 
instruction word (bits 6:0) or the value in the SFR register. 
The 'reg_addr_mux_sel' signal determines which source bus 
to route to 'reg_addr_biis' . The CPU architecture has an 
eight-level stack that enables subroutines to be called via 
the CALL/RETLW instruction pair. Jumps to an absolute 
address can also be done via the GOTO instruction. Figure 3- 
6 below shows the Program Counter Datapath required to 
allow the CALL, RETLW and GOTO instructions to be 
implemented. 



=*.*■ hi rn *nlia I] 




»/ -- ' 



Figure 3-6. Program counter and stack 

As Figure 3-6 shows, the 'pc_mux_sel' input selects the 
source of the data on the 'pc_mux_bus' . The 'pc_ register' 
register is loaded on the rising edge of the clock signal when 
the 'load_pc_reg' signal is asserted. The 'inc_pc' signal is 
used to increment the Program Counter as each instruction is 
executed. The 'inc_stack' and 'dec_stack' signals are used 
to select which stack register will be loaded when the 
'load_stack' signal is asserted. The 'stack_bus' always 
outputs the data in the current stack location (i.e. the data 
previously written to the stack location)[5] . The CPU IP Core 
contains two bidirectional 8-bit I/O and a bidirectional 6-bit 1/ 
O ports .The I/O port bits can individually be programmed as 
input or output via the TRIS instruction. The output data (for 
output ports) is programmed into GPIO registers. Figure 3-7 
below shows the internal architecture of the 'tri_state_port' 
module. 



13 



ACEEE 



ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 201 1 



».i~ /a 






^*J"-*" *■ ft 


} 


\ 




1 


TfJlSfitpiw 




4FlDfl#friir 


**.r*J*t 




1 


■1 









Figure 3-7. I/O port logic 

As can be seen in figure 3-7, the 'alu_biis' input allows 
the TRIS and GPIO 8-bit registers to be loaded from 'alu_bus' . 
A TRIS registers is loaded from 'alu_bus' when a TRIS 
instruction is executed. A GPIO register is loaded when Special 
Function Registers is the write target of an instruction [7] 
(i.e. MOVWF). Setting a bit in the TRIS register puts the 
corresponding output driver in a hi-impedance mode. Clearing 
a bit in the TRIS register puts the contents of the output 
latch on the selected pin. 

Controller Architecture: 

As shown in Figure 3-1, the controller module is 
connected via a number of inputs and outputs to the 
'datapath' module. The controller is responsible for decoding 
the current instruction in the IR (instruction register) and set 
up the datapath control signals so that the data is correctly 
routed from the source to the destination register accordingly 
to the semantics of the instruction. As the instruction in the 
IR changes so does the functionality of the Controller. The 
Controller is essentially a simple state machine that asserts a 
pre-determined number of signals for the various instructions. 
Figure 3-8 below shows the inputs and outputs of the 
'controller' module. 




' -H.l'.. 't 




> -MJilM.ni 


*-K_f"_*7 




■ - >■'• 




>"*_*««_' 




t-.w* 1 .- 




k'M '#* l+J 


*:■■*€_.■ w*_^ 


| 


b*U*fLH 




**Ji' J _ , *l ._ 


"_«_*? 




**J*I<* 




M<J^^ 




■ •-: 'j V 




tt-Jh 




**< ■ ■*■* 




■AJ_*»^*JI* , 




■*I_J**_™ 




j- *ai 1*1 










} 



till 

l'j -He : 



Figure 3-8. Controller 



IV. SIMULATION RESULTS 



Figure 4-1 shows the top level simulation waveforms. This 
figure shows RAM, ALU and instruction decoder waveforms 
also for comparison. These waveforms are explained 
separately in subsequent figures. 




Figure 4-1. Simulation waveforms 

Figure 4-2 shows instruction decoder waveforms. 
Instruction decoding happens based on the instruction format 
described in Chapter 2. MSB 2-bits indicate whether the 
instruction is byte oriented or bit oriented or control 
instruction. For the byte oriented instruction shown in Figure 
4-2, bits 6 down to is the register address for read/write. 




Figure 4-2. Instruction decoder waveforms 




Figure 4-3. ALU waveforms 

As shown in Figure 4-3, alu_mux_bus and w_reg_bus 
are the inputs to the ALU. The output comes out of alu_bus. 
The operation to be performed is decided by the instruction 
coming on ir_reg_biis. The status flags are shown on load_c, 
load_dc and load_z signals. 




Figure 4-4. Program counter waveforms 

As shown in Figure 4-4, the output of program counter 
comes on addr_to_rom. This is changed based on the type 
of instruction executed. By default, it is incremented by ' 1 ' . 



©2011 ACEEE 
DOL01.DEPE.02.02.44 



14 



-fcACEEE 



ACEEE Int. J. on Electrical and Power Engineering, Vol. 02, No. 02, August 201 1 



In case of branch instructions (indicated by branch_ins signal), 
the PC is incremented by value denoted on data_in. Similarly, 
the current address is saved to stack through stack_out (push 
instruction). 




Figure 4-5. ROM waveforms 

Figure 4-5 shows ROM waveforms where the address, chip 
select and corresponding instruction can be seen on data. 

V. FPGAIMPLEMENTATION 

The design is implemented in Spartan 3E FPGA XC3S500E 
on Spartan 3E starter kit. The design occupies 3% of the 
FPGA and correspondingly maximum frequency of operation 
is 233.91MHz. The placed design is shown in Figure 5-1. 
Routed design is shown in Figure 5-2. 



I 




Figure 5-1. Placed design Figure 5-2. Routed design 

The design is implemented in Spartan 3E FPGA and the 
results are verified using chipscope on-chip debugging tool. 
The results obtained are shown in Figure 5-3. 













it S 


Iwllll-^^ 






MUM 


1 


T ? 


m 


m m 


a* 


r. ,^ »v 

*r itlfeH 

' ***■ " ■ ■ ■ 


:■■: 


- 


:■' 


« s 


- 




ni' ii V s. T 


■ 




re 




'■"- W 




1 






lit) 


H 1 






Wa HI C Jttf i BU9 E H4i 3 i&f 5 WK 1 











Figure 5-3. Chipscope results 

For on-chip debugging, some important signals like ROM 
address, instruction, register address, register data input and 
data output are probed. Based on the instruction being 
executed, we can see the signal values. The results are verified 
using chipscope analyzer tool from Xilinx. 

ACKNOWLEDGEMENT 

We would like to thank the Faculty of GNI, KMIT & 
HMIT for their support. We acknowledge Mr satish Babu 
help in the development and testing of our implementation 
"A PIC compatible RISC CPU core Implementation for FPGA 
based Configurable SOC platform for Embedded 
Applications" . 

©2011 ACEEE 
DOr.Ol.DEPE.02.02.44 



CONCLUSION 

We are Implemented RISC based CPU Architecture. This 
CPU uses hardware Architecture and two level Pipelined 
Structures. This Architecture have only 35 reduced 
instruction in it's instruction set and most of the instructions 
only need one machine cycle to be executed. The performance 
of the CPU has been improved by replacing micro program 
with direct Logic Block.This CPU gives good performance 
ratherthan SISC based Architectures. the Simulation Results 
shows the CPU has a good performance. 



REFERENCES 

[1] PIC16C63A/65B/73B/74B Datasheet, Microchip Technology 

Inc,2000. 

[2] Piguet C. Low power design of 8-b embedded cool-RISC 

microcontroller cores. IEEE Journal of Solid-State Circuits, 1997, 

32(7): 1067-1077 

[3] OTP 8-Bit CMOS MCU with EEPROM Data Memory, USA: 

Microchip Technology Inc, 1998 

[4] PIC micro TM Mid-Range MCU Family Reference Manual. 

Microchip Technology Inc. 1997. 1-688. 

[5] Mano, M. M., Computer System Archetecture, Prentice Hall, 

Tsinghua University press, 1997. 

[6] Ma G K, Taylor F J. Multiplier policies for digital signal 

processing. IEEE ASSP Magazine, 1990, 7(1): 6-20 

[7] Hewlett-Packard Company (September 1987). Hewlett-Packard 

Journal 38 

[8] Wayne Wolf, Modern VLSI Design, USA: Printice Hall, 

2001:298-299. 

About Authors 




S Nagakishore Bhavanam is an M.Tech Post- 
Graduate of VLSI-SD from Aurora's Techno- 
logical & Research Institute (ATRI), Depart- 
ment of ECE, JNTUH. He obtained his B.Tech 
from S.V. V.S.N Engineering college, Ongole. He 
has 18 Months of teaching experience. He has 4 
Research Papers, Published in IEEE Xplore, He 
has 3 International journal Publications and has 6 Papers, Published 
in International & National Conferences. His interesting Fields are 
Low Power VLSI, Digital System Design, Sensor Networks, and 
Communications. 

Vasujadevi Midasala is an M.Tech Post- 
Graduate of VLSI from Bandari Institute of 
science and Technology, Department of ECE, 
■ JNTUH. She obtained her B.Tech from S.V.V.S.N 
^^ Engineering college Ongole. She Has 4 Research 
Papers Published in International/ National 
Journals. Her interesting Fields are Low Power 




VLSI, Design for Testability. 

Haritha M is an M.Tech Post- Graduate of VLSI 
from, Aurora's Technological & Research Institute 
(ATRI), Department of ECE, JNTUH. She 
obtained her B.Tech from T.K.R Engineering 
college Hyderabad. Her interesting Fields are 
Microcontrollers & Signals and Signals. 




15 



^VACEEE